|
|
Analysis of Microscopic Images of Cancer Cells
Vičar, Tomáš ; Matula,, Petr (referee) ; Sladoje, Natasa (referee) ; Kolář, Radim (advisor)
Tato disertační práce je zaměřena na analýzu různých forem mikroskopických obrazových dat nádorových buněk (statické 2D snímky, statické 3D obrazy, 2D časosběrné zobrazování živých buněk). Hlavní pozornost je věnována datům získaným koherencí řízeným holografickým mikroskopem, který je relativně novou modalitou schopnou kotrastních záznamů živých buněk bez barvení (label-free) a poskytuje kvantitativní informaci (kvantitativní fázové zobrazení - QPI). V práci je popsán základní postup analýzy těchto snímků a jsou vytvářeny nové metody a zdokonalovány metody pro jednotlivé kroky této analýzy. Největší část práce je věnována segmentaci buněk, kde jsou shrnuty klasické metody i metody založené na hlubokém učení. Jsou také vyvinuty nové metody vhodné právě pro QPI data. Část práce je také věnována segmentaci 3D fluorescenční jader a detekci DNA zlomů pomocí hlubokého učení. Práce se zabývá i dalším zpracování v podobě sledování buněk, extrakce příznaků a následné analýze, kde je detekována buněčná smrt a jsou vytvořeny vhodné interpretovatelné příznaky pro klasifikaci buněčné smrti na apoptickou a lytickou. Celkově tato práce přispívá k rozvoji jednotlivých kroků analýzy obrazu nádorových buněk a odráží současný pokrok v oblasti analýzy obrazu, zejména přístupy hlubokého učení, což je také demonstrováno na několika výzkumných aplikacích.
|
|
|
Tumor cell classification using deep-learning
Majerčík, Jakub ; Kolář, Radim (referee) ; Vičar, Tomáš (advisor)
Classification of microscopic cancer cell images finds its use in a wide variety of biological and medical applications. This work aims to classify two lines of aggressive tumor prostate cells with induced zinc resistance using deep learning methods, and provide an interpretation of occurring classification processes. Dataset consists of more than 750 images, whose acquisition was performed using optical diffraction tomography. This microscopy method allowed for non-invasive cell imaging in their native state. This work shows an implementation of a convolutional neural network, along with methods for visualization of classification processes used to generate localization maps (Grad-CAM and an occlusion-based method). The neural network classifies two prostate cell lines used in study with an accuracy of 98,08% and the aggressive zinc-resistance phenotype with an accuracy of 96,08%. Localization maps and manual segmentation masks of cell borders, nuclei and nucleoli allowed for analysis of sub-celullar regions, which indicates that the decisive region for correct classification is the region of cytoplasm. This is most likely the result of variable vesicle count in cytoplasm, their size, as well as the overall cell size and the morfological structure of their cytoplasmic membrane depending on a given phenotype.
|
|
|
The role of phosphoinositides in macropinocytosis
Hájek, Tomáš ; Doubravská, Lenka (advisor) ; Španielová, Hana (referee)
Macropinocytosis is non-selective actin-dependent type of endocytosis. It is important for the normal physiology of some cell types. However, it is used by intracellular parasites which internalise themselves into host cells in this way and also play a role in the nutritional supply in some type of cancer cells. During macropinocytosis a self-organized subdomain of plasma membrane is separated by a diffusion barrier named macropinocytic cup. RAC1-driven actin polymerization is required for membrane protrusion at the cup periphery, where a narrow ring of the actin nucleating factors is present. In contrast, actin dissociation occurs at the base of the cup due to RAS-controlled formation of phosphatidylinositol trisphosphates (PIP3). During cup closure sequential breakdown of PIP3 to phosphatidylinositol and acquisition of the endosomal identity of the newly formed vesicle is necessary. As a result of tubulation in the early stages of macropinocytosome maturation the vesicle decreases in diameter and stabilizes. At late stages the macropinocytic vesicle may fuse with the lysosome, allowing internalized material to enter this degradative organelle. Throughout the process specific types of phosphatidylinositols are part of the membrane providing signal transduction and membrane identity. These phospholipids...
|
|
|
Analysis of Microscopic Images of Cancer Cells
Vičar, Tomáš ; Matula,, Petr (referee) ; Sladoje, Natasa (referee) ; Kolář, Radim (advisor)
Tato disertační práce je zaměřena na analýzu různých forem mikroskopických obrazových dat nádorových buněk (statické 2D snímky, statické 3D obrazy, 2D časosběrné zobrazování živých buněk). Hlavní pozornost je věnována datům získaným koherencí řízeným holografickým mikroskopem, který je relativně novou modalitou schopnou kotrastních záznamů živých buněk bez barvení (label-free) a poskytuje kvantitativní informaci (kvantitativní fázové zobrazení - QPI). V práci je popsán základní postup analýzy těchto snímků a jsou vytvářeny nové metody a zdokonalovány metody pro jednotlivé kroky této analýzy. Největší část práce je věnována segmentaci buněk, kde jsou shrnuty klasické metody i metody založené na hlubokém učení. Jsou také vyvinuty nové metody vhodné právě pro QPI data. Část práce je také věnována segmentaci 3D fluorescenční jader a detekci DNA zlomů pomocí hlubokého učení. Práce se zabývá i dalším zpracování v podobě sledování buněk, extrakce příznaků a následné analýze, kde je detekována buněčná smrt a jsou vytvořeny vhodné interpretovatelné příznaky pro klasifikaci buněčné smrti na apoptickou a lytickou. Celkově tato práce přispívá k rozvoji jednotlivých kroků analýzy obrazu nádorových buněk a odráží současný pokrok v oblasti analýzy obrazu, zejména přístupy hlubokého učení, což je také demonstrováno na několika výzkumných aplikacích.
|
|
|
Preparation of titanocene complexes with cytostatic properties
Dunlop, David ; Lamač, Martin (advisor) ; Demel, Jan (referee)
Cytotoxicity of titanocene dihalides has been a persistent topic in scientific literature, dating back to the 1970's. The intensively studied titanocene dichloride, which has disappointed in phase II clinical trials on humans, has been substituted by various titanocenes, which bear aryl-substituted cyclopentadienyl ligands. The activity of these complexes can be interestingly further improved via substitution of chloride ligands by fluorides. However, the mechanism of action of these complexes remains somewhat of a mystery, starting with their transport into the cells, stability under physiological conditions, etc. This thesis describes the preparation of several variously substituted titanocene dichlorides and difluorides. These complexes are a part of a larger series, which will help to further the understanding of structural effects on the cytotoxicity of titanocenes. All prepared complexes are characterized via 1 H ,13 C NMR and IR spectroscopy. Complexes containing fluoride ligands have also been characterized via 19 F NMR spectroscopy, melting points, RTG structural analysis (if monocrystals were available) and elemental analysis. Keywords: titanocene dihalides, organometallics, cytotoxicity, cancer cells
|
|
|
Effect of cancer-associated fibroblasts on the survival, proliferation and invasiveness of cancer cells.
Nováková, Gita ; Anděra, Ladislav (advisor) ; Brábek, Jan (referee)
Tumour microenvironment, in addition to cancer cells themselves, represents important structural and functional part of the tumour. Similarly to the normal organs tumour microenvironment comprises several cell types (fibroblasts, immune cells, endothelial cells etc.) and non-cellular components, particularly extracellular matrix. All of them form favourable conditions for the growth, proliferation, protection from the immune system- mediated destruction and nutrition of cancer cells. Cancer associated fibroblasts (CAFs) represent the most abundant cell type of tumour microenvironment. Their origin can be traced to local normal fibroblasts, endothelial cells or epithelial cells and the transition into the CAFs phenotype is influenced with several factors secreted by cancer cells (particularly TGF-β). In contrast to fibroblasts activated during wound healing newly formed cancer associated fibroblasts expressing α-SMA are not subsequently eliminated from the respektive tissue. They persist and produce a number of pro-tumorigenic factors - SDF-1, HGF, IGF-1, IL-6, VEGF, PDGF-C, TGF-β, MMPs etc. CAFs and their secreted factors target several signalling pathways enhancing basic characteristics of the tumour, so called Hallmarks of Cancer. Cancer associated fibroblasts promote proliferation and invasiveness of...
|
|
|
Charakterizace vlivu senescence na indukci a regulaci smrti nádorových buněk
Nováková, Gita ; Anděra, Ladislav (advisor) ; Truksa, Jaroslav (referee)
4 Abstract Senescence is a specific cell state distinquished by cessation of cell division and proliferation and changes in gene expression. Normal cells enter senescence after distinct number of cell divisions or in case of an unrepairable damage. Senescence in cancer cells can be induced by subliminal stress as sublethal treatment with certain drugs. Senescent cancer cells persist in the tissue and may secrete a number of factors and nutrients affecting surrounding cells. Senescence can thus change the response of cancer cells to various apoptogens during cancer therapy. In this study, we focused on the elucidation of presumed differences between normal proliferating and senescent cancer cells in their response to selected apoptogens. Implementing bromodeoxyuridine (BrdU)-mediated replication stress in cancer cells derived from pancreatic (PANC-1) or mesothelioma (H28) tumors, we efficiently forced these cells to acquire senescent phenotype. We document that these senescent cells gain higher resistance to combined TRAIL and homoharringtonine (HHT) treatment and enhance sensitivity to other apoptogens such as FasL, camptothecin and mVES. These cells also showed increased expression of anti-apoptotic protein c-FLIP in senescent cells and changes in the expression of some Bcl-2 family proteins....
|
|
|
Activation and regulation of cell death in senescent cancer cells.
Holíček, Peter ; Anděra, Ladislav (advisor) ; Drbal, Karel (referee)
Cellular senescence is a distinct cell state, characteristic by cessation of cell proliferation and it is accompanied by specific morphological and biochemical alterations. Increasing and persisting incidence of senescence cells has been shown to have detrimental effect on an organism largely contributing to its ageing. Senescent cells also positively support tumour growth and can even stimulate carcinogenic transformation of surrounding cells. Moreover, senescence can be induced even in tumour cells spontaneously or by chemotherapy. Regardless of an initial stimuli and type of cells, there are two main senescence inducing pathways p16/pRb and p53/p21. Both senescent cells as well as senescent cancer cells seems to have modified apoptotic signalling at the level of mitochondria and Bcl-2 family proteins. In this study, we aimed to analyse effect of senescent state as well as pre-senescent (growth arrested state) induced by p16/pRb and p53/p21 signalling pathways on the response of H28 mesothelioma cancer cells-derived clonal cultures to various cell death-inducing stimuli. By inducible expression of p16 and p21 proteins in doxycycline-dependent manner, we forced cells to acquire senescent-like phenotype, which we detailly characterised. Our results showed that senescent-like phenotype, manifests...
|
|
|
Tumor cell classification using deep-learning
Majerčík, Jakub ; Kolář, Radim (referee) ; Vičar, Tomáš (advisor)
Classification of microscopic cancer cell images finds its use in a wide variety of biological and medical applications. This work aims to classify two lines of aggressive tumor prostate cells with induced zinc resistance using deep learning methods, and provide an interpretation of occurring classification processes. Dataset consists of more than 750 images, whose acquisition was performed using optical diffraction tomography. This microscopy method allowed for non-invasive cell imaging in their native state. This work shows an implementation of a convolutional neural network, along with methods for visualization of classification processes used to generate localization maps (Grad-CAM and an occlusion-based method). The neural network classifies two prostate cell lines used in study with an accuracy of 98,08% and the aggressive zinc-resistance phenotype with an accuracy of 96,08%. Localization maps and manual segmentation masks of cell borders, nuclei and nucleoli allowed for analysis of sub-celullar regions, which indicates that the decisive region for correct classification is the region of cytoplasm. This is most likely the result of variable vesicle count in cytoplasm, their size, as well as the overall cell size and the morfological structure of their cytoplasmic membrane depending on a given phenotype.
|
|
|
Preparation of titanocene complexes with cytostatic properties
Dunlop, David ; Lamač, Martin (advisor) ; Demel, Jan (referee)
Cytotoxicity of titanocene dihalides has been a persistent topic in scientific literature, dating back to the 1970's. The intensively studied titanocene dichloride, which has disappointed in phase II clinical trials on humans, has been substituted by various titanocenes, which bear aryl-substituted cyclopentadienyl ligands. The activity of these complexes can be interestingly further improved via substitution of chloride ligands by fluorides. However, the mechanism of action of these complexes remains somewhat of a mystery, starting with their transport into the cells, stability under physiological conditions, etc. This thesis describes the preparation of several variously substituted titanocene dichlorides and difluorides. These complexes are a part of a larger series, which will help to further the understanding of structural effects on the cytotoxicity of titanocenes. All prepared complexes are characterized via 1 H ,13 C NMR and IR spectroscopy. Complexes containing fluoride ligands have also been characterized via 19 F NMR spectroscopy, melting points, RTG structural analysis (if monocrystals were available) and elemental analysis. Keywords: titanocene dihalides, organometallics, cytotoxicity, cancer cells
|
guest ::
